Effect of Variations in the Amounts of P-Glycoprotein (ABCB1), BCRP

Jul 6, 2010 - Effect of Variations in the Amounts of P-Glycoprotein (ABCB1), BCRP (ABCG2) and CYP3A4 along the Human Small Intestine on PBPK Models fo...
2 downloads 10 Views 1MB Size
articles Effect of Variations in the Amounts of P-Glycoprotein (ABCB1), BCRP (ABCG2) and CYP3A4 along the Human Small Intestine on PBPK Models for Predicting Intestinal First Pass Arnaud Bruye`re,†,‡ Xavier Decle`ves,† Francois Bouzom,‡ Kathryn Ball,‡ Catie Marques,† Xavier Treton,§ Marc Pocard,| Patrice Valleur,| Yoram Bouhnik,§ Yves Panis,⊥ Jean-Michel Scherrmann,† and Stephane Mouly*,†,# INSERM U705-CNRS UMR 7157, Faculte´ de Pharmacie, UniVersite´ Paris Descartes, Paris, France, Department of Non-Clinical Modelling, Technologie SerVier, Orle´ans, France, Department of Gastroenterology, Beaujon Hospital, Assistance Publique-Hoˆpitaux de Paris, UniVersite´ Paris VIIsDenis Diderot, Clichy, France, Department of DigestiVe Surgery, Lariboisie`re Hospital, Assistance Publique-Hoˆpitaux de Paris, UniVersite´ Paris VIIsDenis Diderot, Paris, France, Department of DigestiVe Surgery, Beaujon Hospital, Assistance Publique-Hoˆpitaux de Paris, UniVersite´ Paris VIIsDenis Diderot, Clichy, France, and Department of Internal Medicine, Lariboisiere Hospital, Assistance Publique-Hoˆpitaux de Paris, UniVersite´ Paris VIIsDenis Diderot, Paris, France Received January 23, 2010; Revised Manuscript Received July 9, 2010; Accepted July 12, 2010

Abstract: It is difficult to predict the first-pass effect in the human intestine due to a lack of scaling factors for correlating in vitro and in vivo data. We have quantified cytochrome P450/3A4 (CYP3A4) and two ABC transporters, P-glycoprotein (P-gp, ABCB1) and the breast cancer resistant protein BCRP (ABCG2), throughout the human small intestine to determine the scaling factors for predicting clearance from intestinal microsomes and develop a physiologically based pharmacokinetic (PBPK) model. CYP3A4, P-gp and BCRP proteins were quantified by Western blotting and/or enzyme activities in small intestine samples from 19 donors, and mathematical trends of these expressions with intestinal localization were established. Microsome fractions were prepared and used to calculate the amount of microsomal protein per gram of intestine (MPPGI). Our results showed a trend in CYP3A4 expression decrease from the upper to the lower small intestine while P-gp expression is increasing. In contrast, BCRP expression did not vary significantly with position, but varied greatly between individuals. The MPPGI (mg microsomal protein per centimeter intestine) remained constant along the length of the small intestine, at about 1.55 mg/cm. Moreover, intrinsic clearance measured with specific CYP3A4 substrates (midazolam and an in-house Servier drug) and intestinal microsomes was well correlated with the amount of CYP3A4 (R2 > 0.91, p < 0.01). In vivo data were more accurately predicted using PBPK models of blood concentrations of these two substrates based on the segmental distributions of these enzymes and MPPGI determined in this study. Thus, these mathematical trends can be used to predict drug absorption at different intestinal sites and their metabolism can be predicted with the MPPGI. Keywords: Cytochrome P450; human intestine; P-gp; BCRP; PBPK; MPPGI 1596

MOLECULAR PHARMACEUTICS VOL. 7, NO. 5, 1596–1607

10.1021/mp100015x  2010 American Chemical Society

Published on Web 07/06/2010

PBPK Model for Predicting Intestinal First Pass

Introduction Most drugs (over 70%) are administered orally.1 Hence, the small intestine is a vital barrier to their entry into the systemic circulation. The ability to predict the rate and extent of absorption of orally administered compounds is crucial for improved drug development. Factors such as the small intestine permeability, active uptake and efflux transport, drug metabolism, food effects and solubility all influence the amount of a drug entering the portal vein. Previous in Vitro and in ViVo studies have focused on CYP3A-mediated metabolism and P-gp mediated efflux as significant contributors to oral drug bioavailability. The impact of small intestinal first pass on the overall oral bioavailability of drugs has been carefully studied since 19912,3 and is usually included in prediction models. The permeability and metabolism of a drug are key factors for predicting the intestinal first-pass effect, as phase I/phase II drug-metabolizing enzymes and influx/efflux transporters are present all along the human intestinal tract.4 To date, at least seven CYPs have been identified in the human small intestine by mRNA analysis, protein assays and/or enzyme activity.5-13 The ATP-binding cassette (ABC) transporters have also been found throughout the human small intestine, but mainly as their mRNAs (especially the MDR1, BCRP and the MRPs * Correspondence to this author at Hoˆpital Lariboisie`re, Service de Me´decine Interne A, 2 rue Ambroise Pare´, 75010 Paris, France. Tel: 33-1-49 95 81 27. Fax: 33-1-49 95 84 46. E-mail: [email protected]. † Universite´ Paris Descartes. ‡ Technologie Servier. § Department of Gastroenterology, Beaujon Hospital, Assistance Publique-Hoˆpitaux de Paris, Universite´ Paris VIIsDenis Diderot. | Department of Digestive Surgery, Lariboisie`re Hospital, Assistance Publique-Hoˆpitaux de Paris, Universite´ Paris VIIsDenis Diderot. ⊥ Department of Digestive Surgery, Beaujon Hospital, Assistance Publique-Hoˆpitaux de Paris, Universite´ Paris VIIsDenis Diderot. # Department of Internal Medicine, Lariboisiere Hospital, Assistance Publique-Hoˆpitaux de Paris, Universite´ Paris VIIsDenis Diderot. (1) Wilson, C. Gastrointestinal transit and drug absorption. In Oral drug absorption: Prediction and assessmen; Dressman, J. B., Lennerna¨s, H., Ed.; Marcel Dekker: New York, 2000. (2) Kolars, J.; Awni, W.; Merion, R.; Watkins, P. First-pass metabolism of cyclosporine by the gut. Lancet 1991, 14, 1488–1490. (3) Paine, M.; Shen, D.; Kunze, K.; Perkins, J. D.; Marsh, C. L.; McVicar, J. P.; Barr, D. M.; Gillies, B. S.; Thummel, K. E. Firstpass metabolism of midazolam by the human intestine. Clin. Pharmacol. Ther. 1996, 60, 14–24. (4) Benet, L. Z. The drug transporter-metabolism alliance: uncovering and defining the interplay. Mol. Pharmaceutics 2009, 6, 1631– 1643. (5) Paine, M.; Khalighi, M.; Fisher, J.; Shen, D. D.; Kunze, K. L.; Marsh, C. L.; Perkins, J. D.; Thummel, K. E. Characterization of interintestinal and intraintestinal variations in human CYP3Adependent metabolism. J. Pharmacol. Exp. Ther. 1997, 283, 1552– 1562.

articles mRNAs).14-19 Mouly et al. (2003)14 studied four healthy donors and demonstrated that the levels of P-gp increased from the duodenum to the colon. However, there is presently no clear picture of the way in which the abundance of either CYPs or P-gp varies with their location along the human small intestine. Neither is there any clear relationship that can be used to predict intestinal clearance from intrinsic (6) Paine, M.; Ludington, S.; Chen, M.; Stewart, P. W.; Huang, S. M.; Watkins, P. B. Do men and women differ in proximal small intestinal CYP3A or P-glycoprotein expression. Drug Metab. Dispos. 2005, 33, 426–433. (7) Paine, M.; Hart, H.; Ludington, S.; Haining, R. L.; Rettie, A. E.; Zeldin, D. C. The human intestinal cytochrome P450 “pie”. Drug Metab. Dispos. 2006, 34, 880–886. (8) Obach, R.; Zhang, Q.; Dunbar, D.; Kaminsky, L. S. Metabolic characterization of the major human small intestinal cytochrome p450s. Drug Metab. Dispos. 2001, 29, 347–352. (9) Komura, H.; Iwaki, M. Species differences in in Vitro and in ViVo small intestinal metabolism of CYP3A substrates. J. Pharm. Sci. 2007, 97, 1775–1800. (10) Madani, S.; Paine, M.; Lewis, L.; Thummel, K. E.; Shen, D. D. Comparison of CYP2D6 content and metoprolol oxidation between microsomes isolated from human livers and small intestines. Pharm. Res. 1999, 16, 1199–1205. (11) Lown, K.; Kolars, J.; Thummel, K.; Barnett, J. L.; Kunze, K. L.; Wrighton, S. A.; Watkins, P. B. Interpatient heterogeneity in expression of CYP3A4 and CYP3A5 in small bowel. Lack of prediction by the erythromycin breath test. Drug Metab. Dispos. 1994, 22, 947–955. (12) Zhang, Q.; Dunbar, D.; Ostrowska, A.; Zeisloft, S.; Yang, J.; Kaminsky, L. S. Characterization of human small intestinal cytochromes P-450. Drug Metab. Dispos. 1999, 27, 804–809. (13) Zeldin, D.; Foley, J.; Goldsworthy, S.; Cook, M. E.; Boyle, J. E.; Ma, J.; Moomaw, C. R.; Tomer, K. B.; Steenbergen, C.; Wu, S. CYP2J subfamily cytochrome P450s in the gastrointestinal tract: expression, localization, and potential functional significance. Mol. Pharmacol. 1997, 51, 931–943. (14) Mouly, S.; Paine, M. P-glycoprotein increases from proximal to distal regions of human small intestine. Pharm. Res. 2003, 20, 1595–1599. (15) Taipalensuu, J.; To¨rnblom, H.; Lindberg, G.; Einarsson, C.; Sjo¨qvist, F.; Melhus, H.; Garberg, P.; Sjo¨stro¨m, B.; Lundgren, B.; Artursson, P. Correlation of gene expression of ten drug efflux proteins of the ATP-binding cassette transporter family in normal human jejunum and in human intestinal epithelial Caco-2 cell monolayers. J. Pharmacol. Exp. Ther. 2001, 299, 164–170. (16) Chan, L.; Lowes, S.; Hirst, B. The ABCs of drug transport in intestine and liver: efflux proteins limiting drug absorption and bioavailability. Eur. J. Pharm. Sci. 2004, 21, 25–51. (17) Zimmermann, C.; Gutmann, H.; Hruz, P.; Gutzwiller, J. P.; Beglinger, C.; Drewe, J. Mapping of multidrug resistance gene 1 and multidrug resistance-associated protein isoform 1 to 5 mRNA expression along the human intestinal tract. Drug Metab. Dispos. 2005, 33, 219–224. (18) Hilgendorf, C.; Ahlin, G.; Seithel, A.; Artursson, P.; Ungell, A. L.; Karlsson, J. Expression of thirty-six drug transporter genes in human intestine, liver, kidney, and organotypic cell lines. Drug Metab. Dispos. 2007, 35, 1333–1340. (19) Seithel, A.; Karlsson, J.; Hilgendorf, C.; Bjo¨rquist, A.; Ungell, A. L. Variability in mRNA expression of ABC- and SLCtransporters in human intestinal cells: comparison between human segments and Caco-2 cells. Eur. J. Pharm. Sci. 2006, 28, 291– 299. VOL. 7, NO. 5 MOLECULAR PHARMACEUTICS 1597

articles clearance due to a lack of scaling factors for the human small intestine. A scaling factor is known for the liver, and hepatic data have frequently been used to predict intestinal first pass, based on the relative amounts of CYP3A4 in the liver and intestine. But this approach is far from physiological, despite some predictions being similar to observed in ViVo data.20 Many software tools are presently available for predicting the first-pass effect and bioavailability of drugs from in Vitro data. Although prediction models are becoming more and more reliable, particularly the physiologically based pharmacokinetic models (PBPK), they often fail to accurately predict the extent of intestinal metabolism, mainly due to a lack of data on the amount of microsomal protein per gram of intestine (MPPGI) and the concentration of CYP in each of the intestinal segments. PBPK models need to be refined to make them more pertinent. Studies by Paine et al. (2006)7 and Badhan et al. (2009)20 have estimated the MPPGI to be 2 mg per gram of gut. This factor is commonly used in the Simcyp PBPK model and considers the intestine to be a homogeneous metabolic organ, with different amount of CYP3A4 along the gastrointestinal tract. However, the intestine is not a homogeneous tissue. Hence, factors such as the segmental distribution of intestinal enzymes and MPPGI need to be accurately determined for each segment of the small intestine. Although determining these relationships may require several human samples, the resulting data could lead to better prediction of the impact of the intestinal first-pass effect. For example, it could allow the prediction of how orally administered drug are metabolized and thus enable doses to be adjusted to specific pathological conditions. This study focuses on CYP3A4 and P-gp because they are the main proteins involved in the metabolism of over 50% of currently marketed drugs.22 CYP3A4 also accounts for almost 80% of total intestinal CYPs.7 In addition, their substrate specificities overlap substantially.23,24 Thus intestinal CYP3A4 and P-gp may regulate the availability of oral drugs.24 Lastly, the substrate specificity of BCRP overlaps to a considerable, but varying, degree that of P-gp.25,26 We therefore determined the distributions of both proteins all along the intestinal tract. (20) Badhan, R.; Penny, J.; Galetin, A.; Houston, B. Methodology for development of a physiological model incorporating CYP3A and P-Glycoprotein for the prediction of intestinal drug absorption. J. Pharm. Sci. 2008, 98, 2180–2197. (21) Hakooz, N.; Ito, K.; Rawden, H.; Gill, H.; Lemmers, L.; Boobis, A. R.; Edwards, R. J.; Carlile, D. J.; Lake, B. G.; Houston, J. B. Determination of a human hepatic microsomal scaling factor for predicting in vivo drug clearance. Pharm. Res. 2006, 23, 533– 539. (22) Zhou, S. Potential strategies for minimizing mechanism-based inhibition of cytochrome P450 3A4. Curr. Pharm. Des. 2008, 14, 990–1000. (23) Wacher, V.; Silverman, J.; Zhang, Y.; Benet, L. Z. Role of P-glycoprotein and cytochrome P450 3A in limiting oral absorption of peptides and peptidomimetics. J. Pharm. Sci. 1998, 87, 1322–1330. (24) Watkins, P. The barrier function of CYP3A4 and P-glycoprotein in the small bowel. AdV. Drug DeliVery 1997, 27, 161–170. 1598

MOLECULAR PHARMACEUTICS VOL. 7, NO. 5

Bruyère et al. The resulting data were combined to establish a mathematical trend between the amounts of CYP3A4, BCRP and P-gp and the distance from the stomach to the colon. We also measured the enzymatic activities of CYP3A4 in microsomal protein extracts taken along the whole length of the small intestine using midazolam and a Servier company in-house drug. These data were used to determine the MPPGI for each section of intestine, as previously described for the liver.21 Finally, we challenged our in Vitro results and scaling factors with a physiologically based pharmacokinetic (PBPK) model to predict the plasma/blood concentration profiles of oral administered midazolam and the Servier drug. This model uses the newly determined MPPGIs and the segmental expression of CYP3A4 and P-gp.

Patients, Materials and Methods Chemicals. Midazolam (MDZ), 1′-hydroxy midazolam (1′-OH MDZ), protease inhibitors and organic solvents were purchased from Sigma-Aldrich (St. Louis, MO). The Servier drug (S drug), a specific CYP3A4 substrate, and its hydroxylated primary metabolite were a gift from Technologie Servier (Orleans, France). Midazolam and the Servier company in-house drug (S drug) parameters are detailed in Perdaems et al.27 Patients and Tissue Procurement. The collection of human tissues (n ) 9) and intestinal biopsies (n ) 10) and their use were approved by the local Ethics Committee of ParissHopital Saint Louis. All patients gave written informed consent prior to tissue collection. Duodenal and ileal pinch biopsies were obtained from 10 patients, mean age 58 ( 16 years (range: 33-79 years), undergoing upper and/or lower endoscopy for gastric disorders (n ) 2), diarrhea (n ) 3), intestinal transplant (n ) 2), or polypectomy (n ) 3) (Table 1). Duodenal, jejunal and ileal samples were collected from 9 patients, mean age 57.5 ( 17 years (range: 36 to 82) during surgery for pancreatic (n ) 3), intestinal (n ) 2) or colon cancer (n ) 4) (Table 1). The “healthy” part of the sample was used; it was taken as far as possible from the tumor, as determined by microscopic examination during surgery. Surgeons evaluated the distance of the sample from the stomach according to the excision protocol. A pool of ileum biopsies was formed from three patients undergoing lower endoscopy for suspected polyps, but all of them were considered healthy, as all donors included in the study. These (25) Doyle, L.; Ross, D. Multidrug resistance mediated by the breast cancer resistance protein BCRP (ABCG2). Oncogene 2003, 22, 7340–7358. (26) Haimeur, A.; Conseil, G.; Deeley, R.; Cole, S. P. The MRP-related and BCRP/ABCG2 multidrug resistance proteins: biology, substrate specificity and regulation. Curr. Drug Metab. 2004, 5, 21– 53. (27) Perdaems, N.; Blasco, H.; Vinson, C.; Chenel, M.; Whalley, S.; Cazade, F.; Bouzom, F. Predictions of Metabolic Drug-Drug Interactions Using Physiologically Based Modelling: Two Cytochrome P450 3A4 Substrates Coadministered with Ketoconazole or Verapamil. Clin. Pharmacokinet. 2010, 49, 239–258.

articles

PBPK Model for Predicting Intestinal First Pass Table 1. Intestine Samplesa analysis

a

sample

location

type

disease

CYP3A4

P-gp

BCRP

D1 D2 D3 D4 D5 J1 J2 J3 J4 J5 I1 I2 I3 I4 I5 I6 C1 C2 C3

duodenum duodenum duodenum duodenum duodenum proximal jejunum proximal jejunum medium jejunum medium jejunum medium jejunum ileum ileum ileum ileum ileum ileum colon colon colon

biopsy biopsy surgery surgery surgery biopsy surgery surgery biopsy biopsy biopsy surgery surgery surgery surgery biopsy biopsy biopsy biopsy

gastric disorder gastric disorder pancreatic cancer pancreatic cancer pancreatic cancer gastric disorder small intestinal cancer small intestinal cancer intestinal transplantation intestinal transplantation small intestinal cancer small intestinal cancer small intestinal cancer colon cancer colon cancer small intestinal cancer chronic constipation chronic diarrhea polypectomy

WB, microsome WB, microsome microsome WB WB WB WB, microsome microsome WB WB WB microsome WB, microsome microsome microsome WB WB WB WB

WB WB NA WB WB WB WB NA WB WB WB NA WB NA NA WB WB WB WB

WB WB NA WB WB WB WB NA WB WB WB NA WB NA NA WB WB WB WB

NA: Not applicable. WB: Western blot. Microsome: microsomal preparation.

patients were not included in individual analyses. As surgery samples and biopsies were considered as healthy, all samples were comparable. No difference in the protein expression was observed between biopsies and surgery samples, whatever the suspected pathology. Yet, expressions of P-gp, CYP3A4 and BCRP were determined in patients, suggesting that these results have to be interpreted with caution before extrapolating CYP3A4, P-gp and BCRP expression from our patients to patients without intestinal disease. Preparation of Intestinal Microsomes from Surgical Samples. Intestinal microsomes were prepared using the method developed for rats28 and adapted for human. Each intestine sample was cooled to 4 °C and immediately washed with 30 mL of 0.9% NaCl, 0.5 mM DTT and a cocktail of protease inhibitors (PI). The sample was placed in a Petri dish and the mucosa scraped off with two microscope slides into phosphate-buffered saline (PBS) solution without calcium and magnesium, 1.5 mM EDTA, 3 UI/mL heparin, 0.5 mM DTT and PI. This tissue was immediately incubated on ice for 20 min in 35 mL of the same solution with horizontal agitation. It was then centrifuged at 900g for 10 min, and the resulting pellet was homogenized with a motor-driven Potter-Elvehjem homogenizer in PBS containing 1 mM EDTA, 5 mM histidine, 0.25 mM sucrose and PI (with or without 20% glycerol). This centrifugation/homogenization was repeated. The final pellet was homogenized with the previous buffer (3 mL per gram cells) in a Potter-Elvehjem homogenizer. The resulting homogenate was centrifuged at

10000g for 15 min, and the supernatant was ultracentrifuged for 90 min at 100000g. The resulting microsomal pellet was homogenized with a Potter-Elvejhem homogenizer in TrisHCl, pH 7.4 containing PI and stored at -80 °C. We determined the MPPGI from the weight and length of the tissue sample, the weight of the scraped-off mucosa and total microsomal protein. The concentrations of microsomal protein were measured using the BC assay (Sigma-Aldrich, St. Louis, MO), and the amount of CYP by the method of Matsubara et al. (1976)29 and Omura and Sato (1964).30 Determination of CYP3A4 Activity in Intestinal Microsomes. Intestinal microsomes (1 mg/mL) in 0.1 mM Tris-HCl, pH 7.4 and 5 mM MgCl2 were incubated at 37 °C with MDZ (0.5 µM) for 30 min; while other microsome samples (2 mg/mL) were incubated with S drug (5 µM) under the same conditions. After a 15 min preincubation, reactions were initiated by adding 1 mM NADPH; they were stopped by adding an equal volume of methanol. Samples were taken at 5 times to obtain linear kinetics. MDZ, S drug and their respective metabolites were separated by high-performance liquid chromatography (HPLC) with an Interchim Symmetry C18 (4.6 × 250 mm) column. The elution gradients were as follows: water/ammonium formate (20 mM) and acetonitrile from 70/30 to 10/90 v/v over 20 min with a plateau from 4 to 10 min at 55/45 v/v. MDZ and S drug were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/ MS); MDZ at transitions 326 > 291.2 m/z and S drug at 469.3

(28) Bruyere, A.; Decleves, X.; Bouzom, F.; Proust, L.; Martinet, M.; Walther, B.; Parmentier, Y. Development of an optimized procedure for preparation of rat intestinal microsomes: comparison of hepatic and intestinal microsomal cytochrome P450 enzyme activities in two rat strains. Xenobiotica 2009, 39, 22–32.

(29) Matsubara, T.; Koike, M.; Touchi, A.; Tochino, Y.; Sugeno, K. Quantitative determination of cytochrome P450 in rat liver homogenates. Anal. Biochem. 1976, 75, 596–603. (30) Omura, T.; Sato, R. The carbon monoxyde-binding pigment of liver microsomes. J. Biol. Chem. 1964, 239, 2370–2378. VOL. 7, NO. 5 MOLECULAR PHARMACEUTICS 1599

articles > 177 m/z (positive electrospray ionization). Both activities were quantified from the formation of their respective metabolites normalized to an internal standard. The limit of quantification of MDZ was below 0.001 µM, and that of S drug was below 0.01 µM. Calibration curves were linear over the range studied (5 intermediate concentrations from 0.001 µM to 1 µM for MDZ and from 0.01 to 10 µM for S drug with R2 > 0.99, and CV were less than 5%). An internal standard was used to control injected volume during LC-MS/ MS run (injected volume of 100 µL ( 0.5 µL). Intrinsic clearance (Clint ) Vmax/Km) was calculated from the slope of the plot of substrate disappearance against incubation time (initial conditions). The concentrations used were well below the Km and were tested to ensure that kinetics were firstorder. The slope of substrate disappearance was divided by the protein concentration to obtain an intrinsic clearance in mL/min/g protein. Western Blotting. The amounts of CYP3A4, P-gp and BCRP proteins along the small intestine were determined in homogenates prepared from intestinal biopsies and surgical samples (from 5 mg of total tissue). The amount of CYP3A4 in the microsomes prepared from each section of the intestine was also measured. Proteins were separated by electrophoresis on 12% sodium dodecyl sulfate polyacrylamide gels (SDS-PAGE) at 100 mV for 1 h (gels for P-gp were 8% SDS-PAGE), and electrotransferred to nitrocellulose membranes at 80 mV for 2 h (samples were analyzed in duplicate on two separate Western blot membranes at two different days). Nonspecific binding sites were blocked by incubation overnight at 4 °C with 5% (w/v) non-fat dried milk in a buffer (TBS-T) containing 10 mM Tris-HCl, pH 7.5, 200 mM NaCl and 0.1% (v/v) Tween 20. The membranes were incubated for 2 h with rabbit or mouse monoclonal antibodies against human CYP3A4 (diluted 1:2000) (Tebu-bio, Le Perray-en-Yvelines, France), P-gp (1:200) (C219 Alexis biochemicals, San Diego, CA), BCRP (1:200) (BXP21 Abcam, Paris, France) and mouse polyclonal antibodies against human CYP3A5 (1:2000) (Tebu-bio, Le Perray-enYvelines, France). They were washed with TBS-T and then incubated with enhanced chemiluminescence (ECL) antirabbit or anti-mouse secondary horseradish peroxidasecoupled IgG antibody (diluted 1/10000, except for BCRP where antibodies were diluted 1/1000) for one hour (Amersham Biosciences Europe GmbH, Orsay, France) and then washed three times with TBS-T. Membranes were exposed to Amersham ECL blotting detection reagent (ECL; Amersham Biosciences Europe GmbH, Orsay, France) for 5 min. Signals were revealed with the Bio-Rad ChemiDocTM XRS imaging device and the Quantity One 1-D software (BioRad Laboratories, Munich, Germany). The absolute amount of CYP3A4 was determined using supersomes (BD Biosciences, Woburn, MA) as standard (0.025-25 pmol/mg). For P-gp and BCRP, the housekeeping protein β-actin was used to normalize samples and avoid interday variations. A positive control (human liver sample) was added on each Western blot. Western blots were first validated using purified recombinant protein for P-gp and BCRP assays (Abnova, 1600

MOLECULAR PHARMACEUTICS VOL. 7, NO. 5

Bruyère et al.

Figure 1. Representation of the PBPK model build to

predict in vivo data of midazolam and S drug.

Heidelberg, Germany) and with both supersomes expressed CYP3A4 (BD Gentest, Le Pont de Claix, France) and liver sample for CYP3A4. The reliability of this semiquantitative analysis of protein expression was proven by the linearity of the calibration curve for CYP3A4 (from 0.02 pmol/mg to 20 pmol/mg) and the ratio of the intensities of P-gp or BCRP (from 10 µg to 100 µg per well) divided by that of β-actin. The relationship between the amount of CYP3A4 or the relative abundance of P-gp and BCRP and the position along the human intestine was studied using Sigmaplot software (Systat Software, San Jose, CA). The abundance of P-gp and BCRP was expressed as a relative value with the reference being a pool of ileal tissue from three patients. Therefore, before comparison with the pool, each individual data was normalized with the housekeeping protein β-actin. The pylorus was considered to be the start of the intestine (0 cm), and the jejunum was assumed to start 25 cm after the pylorus. The best mathematical function (linear, exponential, logarithmic and inverse) was selected on the basis of the R2. PBPK Modeling. The model was built with ASCLXtrem software (The AEgis Technologies Group Inc., Huntsville, AL). The parameters of each molecule and the equations used in the model have been described by Perdaems et al.27 Both drugs were studied in solution, and as a consequence no dissolution equation was included in the model. Briefly, the absorption, distribution, the liver and intestinal first-pass metabolism and elimination of a drug are predicted by the software (Figure 1). The model differs from that of Perdams et al.27 in that the gastrointestinal tract was divided into seven sections (1 for the stomach, 1 for the duodenum, 2 for the jejunum, 2 for the ileum and 1 for the colon). Simulations were performed using a model man (70 kg body weight; cardiac blood flow 363 L/h). Organ volumes were expressed as a percentage of body weight and organ blood flows as a percentage of the cardiac blood flow.27–31 32 The assumptions for the model were as follows: (a) intercompartmental transport occurred via the blood, (b) drug concentrations in (31) Davies, B.; Morris, T. Physiological parameters in laboratory animals and humans. Pharm. Res. 1993, 10, 1093–1095. (32) Brown, R. P.; Delp, M. D.; Lindstedt, S. L.; Rhomberg, L. R.; Beliles, R. P. Physiological parameter values for physiologically based pharmacokinetic models. Toxicol. Ind. Health 1997, 13, 407–484.

articles

PBPK Model for Predicting Intestinal First Pass effluent blood and blood within tissues were equal, (c) there was instantaneous equilibrium between tissue and blood within the tissue, and (d) only unbound drugs were eliminated. The model was based on a well-stirred model for all organs and intestinal sections. The metabolism was evaluated from microsomes according to the following equations: Clliver )

Clint(liver) × MPPGL × weightliver × Qliver × fu Qliver + fu × Clint(liver) × MPPGL × weightliver (1)

and Clduodenum ) Clint(duodenum) × MPPGIduodenum × weightduodenum × Qduodenum × fu Qduodenum + fu × Clint(duodenum) × MPPGIduodenum × weightduodenum (2)

and Clintestine ) Clduodenum + Cljejunum + Clileum + Clcolon (3) where MPPGL is microsomal protein per gram of liver, MPPGI is the amount of microsomal protein per gram of intestine, weightliver and weightintestine are the weights of these organs, Q is the blood flow within the tissue, fu is the unbound fraction (1 in the intestine), Clint(liver) is the intrinsic clearance measured on liver microsomes and Clint(duodenum) is the intrinsic clearance measured on duodenum microsomes.

Results Quantification of CYP3A4, P-gp and BCRP in Human Intestinal Samples. The amounts of CYP3A4 in biopsies and surgical samples were determined using 5 mg of intestinal tissue. The mean amount of CYP3A4 in 4 samples of duodenum was 1.74 ( 0.83 pmol/mg total protein (Table 2), showing the great interindividual variations in the concentration of CYP3A4 (CV ) 47%). The concentration of CYP3A4 in the jejunum was 3-fold lower than that in the duodenum; the mean concentration was 0.53 ( 0.42 pmol/mg (CV ) 77%) total protein. The concentration of CYP3A4 in the proximal jejunum was close to that in the duodenum and roughly 3 times higher than in the medium jejunum; it was below the limit of quantification (